ROOT HAIRS AND THE RHIZOBIUM/LEGUME SYMBIOSIS

The root hairs of many agriculturally-important legumes are invaded by the symbiotic soil-dwelling bacterium Rhizobium, and there has been recent concentration on root hair cell interaction with lipo-oligosaccharides (Nod-factors) (eg Heidstra et al. 1994; Relic et al. 1994) that are excreted from rhizobia and which are thought to be the compounds involved in early recognition and in signal transduction (Spaink, 1992). Rhizobia and Nod-factors are able to perturb the normal growth processes of legume root hairs (modeled in Ridge, 1993a), and they are therefore useful tools for the study of root hair growth.

When rhizobia are inoculated onto their host plants, they cause distortions, branching, and curling of root hairs. It is within the curled root hair that rhizobia are able to invade the cell wall matrix, and at the plasma membrane interface, cause the hair to cease growth and instead deposit wall material at the invasion site. The gradual build up of this wall material becomes an infection 'thread' (within which the rhizobia are confined) that continues to grow towards the base of the root hair cell, eventually fusing with the basal wall, and there is then a repetition of the initial invasion steps as the rhizobia invade the cortex cells. Ahead of these events, cell division occurs in the cortex of the root, and the infection thread invasion and subsequent release of rhizobia forms the nitrogen fixing nodule.

It is clear that in the very initial interactions with root hairs, rhizobia are able to disturb normal polarity of root hair growth, and in studies on how polarity is maintained, rhizobia should provide useful information. For example, how are calcium gradients and membrane potentials disturbed during this initial interaction? Ehrhardt et al. (1992) have already shown that Nod factors can depolarise alfalfa root hair membrane potential, but no other research in this area has been forthcoming.

One of the most prominent features of the initial invasion is the leading role of the root hair nucleus, which precedes the tip of the infection thread at a set distance, in much the same way as it 'follows' the tip of normal root hairs as they grow. Indeed, it was this phenomenon that was the impetus for the study by Lloyd et al. (1987) on the role of the cytoskeleton in nuclear migration of the legume root hair.

How rhizobia are able to 'sequester' the tip growth machinery of the root hair and use it to form an infection thread is a very interesting biological puzzle. In a model of Rhizobium infection (Ridge, 1993a), it is suggested that the initial action of Rhizobium is to perturb and then de-couple the cytoskeletal link that is connected to the root hair tip. The initiation of infection threads may occur when tip growth is thus sequestered by this appropriation of the cytoskeleton, and microfilaments are induced to unload their vesicle passengers at that site by temporary depolymerisation as rhizobia penetrate the root hair cell wall. Rhizobium may also cause the switching off of the cell wall 'plasticising' effect of normal tip growth (involvement of expansins?), but not the deposition of cell wall precursors, allowing the inward growth of an infection thread. As infection thread growth commences, the cytoskeleton is coupled to the tip of the thread, establishing an important link with the nucleus, similar to the link in normal root hair tip growth.

One basis for the idea that rhizobia are able to induce vesicle unloading at a site other than at the tip comes from a study on the effects of cytochalasin D. In a freeze-substitution study, Ridge (1990) showed that when microfilaments are fragmented, vesicles (that are presumably being transported by cytoplasmic streaming) immediately migrate to the closest cell wall, fusing with the membrane and depositing their contents. The evidence for this is the occurrence of randomly deposited lumps of cell wall at sites where organelles crowd together after cytochalasin treatment (Ridge, 1990). This work has not been confirmed elsewhere. However, it shows that there probably aren't any factors that control the deposition of vesicle contents specifically at the root hair tip.

Though it has yet to be demonstrated clearly, there is every reason to believe that the cytoskeleton is involved in the nucleus leading the infection thread to the base of the hair. If this is so, then what factors are rhizobia unlinking to be able to takeover the tip growth machinery? This sequestration of the tip growth machinery by rhizobia implies that there are cytoskeletal-associated proteins that help to maintain a continuous link to the growing tip, and that these are first unlinked from the tip of the hair and then re-linked to the site of the growing infection thread.

Rhizobia and Nod-factors are thus an exciting tool for the biologist interested in understanding the cell biology of root hairs.